Status of the experiments on measurement of the Newtonian

advertisement
c Pleiades Publishing, Ltd., 2008.
ISSN 0202-2893, Gravitation and Cosmology, 2008, Vol. 14, No. 4, pp. 368–375. Status of the Experiments on Measurement of the Newtonian
Gravitational Constant
V. K. Milyukov1* , Jun Luo2 , Chen Tao1 , and A. P. Mironov1
1
Sternberg Astronomical Institute, Moscow State University,
Universitetskii pr. 13, Moscow 119192, Russia
2
Department of Physics, Huazhong University of Science
and Technology, Wuhan 430074, People’s Republic of China
Received February 12, 2008
Abstract—Due to the weakness of gravity, the accuracy of the Newtonian gravitational constant G is
essentially below the accuracy of other fundamental constants. The current value of G, recommended
by CODATA in 2006, based on all results available at the end of 2006, is G = (6.67428 ± 0.00067) ×
10−11 m3 kg−1 s−2 with a relative error of 100 ppm. The accuracy of the best experimental results is
15–40 ppm, although the scatter of the results is large enough. Therefore new experiments at a level of
accuracy of 10–30 ppm are rather topical. One of the problems of improving accuracy of G is a precision
measurement of the period of eigen oscillations of a torsion balance. The nonlinear behavior of the torsion
balance with five degrees of freedom has been studied. It was shown that swing modes are excited by
the acting environmental noise. A coupling of the swing modes to the torsional mode has been revealed.
Methods of suppressing the effect of mode couplings have been considered.
PACS numbers: 04.80.Cc, 02.60.Cb
DOI: 10.1134/S0202289308040130
1. INTRODUCTION
The Newtonian gravitational constant G together
with Planck’s constant and the speed of light c are
the fundamental constants of nature which represent
fundamental limits: c is the maximum speed, is the
minimum angular momentum and G is the gravitational radius of unit mass (the minimum radius of a
sphere for relativistic gravitational collapse).
While the absolute values of the fundamental constants c and are known with high accuracy and
their “constancy” is not put to doubt, the situation
with the gravitational constant G is absolutely different. Due to the weakness and nonshieldability of
the gravitational interaction, the accuracy of experimental determination of G is essentially below that of
other fundamental constants. Measurements of the
gravitational constant are connected with absolute
measurements of three physical values: time, mass
and length, and consequently it is necessary to perform absolute measurements at high technology level
in order to have a reliable estimation of G.
The first device for measurement of mutual gravitational attraction of small laboratory bodies, the
horizontal torsion balance, has been made at the end
*
E-mail: milyukov@sai.msu.ru
of the 18th century by Henry Cavendish. Hundred
years after Newton’s discovery of the law of gravity,
in 1797–1798, he performed an experiment and determined the value of the gravitational constant, G =
(6.67 ± 0.07) × 10−11 m3 kg−1 s−2 , with a relative
uncertainty of 104 ppm (part per million, i.e., ×10−6 ).
Cavendish also determined the mass and mean density of the Earth. The significance of the Cavendish’s
experiment was not only restricted to determination
of the G value. The main thing is that he has proved
the validity of the universal law of gravitation for small
laboratory bodies.
The modern experimental facilities for measurement of the gravitational constant are complicated
devices performed at high technology level, but their
main part is also a horizontal torsion balance. After
2000, five new results on measurement of G with
a relative error less than 50 ppm have been published. However, these results also did not cover
each other within confidential intervals. In 2006,
the Committee on Data for Science and Technology (CODATA) recommended for the Newtonian
gravitational constant a value of G = (6.67428 ±
0.00067) × 10−11 m3 kg−1 s−2 , with an uncertainty
of 100 ppm. This new value of G is based on the data
accessible by the end of 2006.
368
STATUS OF THE EXPERIMENTS
Since Cavendish’s first laboratory measurement
over 200 years ago, the reduction in uncertainty of G
has been only two orders of magnitude. The progress
in measurement of G is occurring slowly enough:
the error value decreases by approximately a factor
of 10 per century, and the knowledge of the absolute
value of G is still rather poor. New experiments on
measurement of the Newtonian gravitational constant at the accuracy level of 10–30 ppm are topical
and desirable.
369
Suspension wire
Attracting mass
Test mass
Light
Mirror
2. THE PRINCIPLE OF CAVENDISH-TYPE
EXPERIMENT
The torsion balance, or torsion pendulum, is a
traditional tool for performing high-precision gravitational experiments. The principle of Cavendishtype experiments on measurement of the Newtonian
gravitational constant is as follows. A bar with two
equal test masses is suspended on a thin torsion wire
in the gravitational field of big source (attracting)
masses (the bar and the torsion wire form the torsion
balance). Due to the gravitational interaction of the
test and source masses, the torsion balance is twisted
by some angle which is monitored by an optical lever
(Fig. 1).
In most of modern experiments, the gravitational
constant is determined by the so-called time-ofswing method. The time-of-swing method, developed by Heyl [1], is based on detecting the change in
the eigen oscillation frequency of the torsion balance
in the gravitational field of the source masses in two
different configurations, referred to as “near” and
“far” positions. Since in both configurations the
torsion balance oscillates in the gravitational field
produced by the source masses, the gravitational
torsion constant (the “gravitational rigidity”) Kg (ϕ)
is added to the elastic torsion constant of the wire
Ke . As a result, the squared oscillation frequency
becomes
Ke + Kg (ϕ)
,
(1)
ω 2 (ϕ) =
J
where J is the of moment of inertia of the torsion bar
relative to the wire axis; ϕ is the angle between the
pendulum axis and that of the source masses.
The gravitational rigidity is a derivative of the moment of interaction force between the source masses
and the torsion balance,
∂Γ
= GCg (ϕ),
(2)
Kg (ϕ) =
∂ϕ
where Cg (ϕ) is the coupling coefficient determined by
the geometric dimensions of the torsion balance and
attracting masses.
GRAVITATION AND COSMOLOGY Vol. 14
No. 4
2008
Attracting mass
Test mass
Scale
Fig. 1. The principle of Cavendish-type experiments.
The frequency ω is measured in the process of the
experiment. The moment of inertia J and the coupling
constant Cg (ϕ) are calculated on the basis of weighing and measuring of geometrical dimensions of the
torsion balance and source masses. Hence in Eq. (1)
there are two unknown parameters (if we substitute
Eq. (2) to Eq. (1)): the elastic torsion constant Ke and
the gravitational constant G. Therefore to determine
G it is necessary to measure the frequency ω at least
for two different configurations of the source masses,
the “near” and “far” positions. So, the gravitational
constant G can be obtained as
∆ω 2
(ω 2 )1 − (ω 2 )2
=J
,
(3)
G=J
(Cg )1 − (Cg )2
∆Cg
where the indices 1 and 2 indicate two different positions of the attracting masses.
3. THE MODERN HISTORY
OF G DETERMINATION
The modern history of G determination covers 30
to 35 years and has started from three experiments
performed in the 70-s of the last century. These
were the experiment of Observatoire de Recherches
de la Meteorologie Nationale (France), reported
in 1972 [2], the experiment of Sternberg Astronomical Institute of Moscow University, reported in
1979 [3], and the experiment of the National Bureau
of Standards (USA), reported in 1982 [4]. The
CODATA system of values of fundamental constants
of 1986 contained the G value with a relative uncertainty of 128 ppm, which was mainly based on the
value obtained by Luther and Towler [4], but with
370
MILYUKOV et al.
The world’s best experiments on measurement of G and CODATA values
Authors, year of publication
G, ×10−11
STD, ×10−11
m3 kg−1 s−2
m3 kg−1 s−2
ppm
Ponticis, 1972
[2]
6.6714
0.0006
90
Sagitov, Milyukov, et al., 1979
[3]
6.6745
0.0008
120
Luther and Towler, 1982
[4]
6.6726
0.0005
75
CODATA, 1986
[5]
6.67259
0.00085
Michaelis et al., 1995
[6]
6.7154
0.0006
90
Karagioz, Izmailov, 1996
[7]
6.6729
0.0005
75
Bagley, Luther, 1997
[8]
6.6740
0.0007
105
CODATA, 1998
[11]
6.673
0.010
1500
Jun Luo et al., 1999
[10]
6.6699
0.0007
105
Fitzgerald, Armstrong, 1999
[9]
6.6742
0.0007
105
Gundlach, Merkowich, 2000
[12]
6.674215
0.000092
14
Quinn, Speake et al., 2001
[13]
6.67559
0.00027
41
Schlamminger et al., 2002
[14]
6.67407
0.00022
33
CODATA, 2002
[15]
6.6742
0.0010
150
Armstrong, Fitzgerald, 2003
[16]
6.67387
0.00027
41
Schlamminger et al., 2006
[17]
6.67425
0.00010
16
6.67428
0.00067
100
CODATA, 2006
its uncertainty doubled, which reflects the fact that,
historically, measurements of G have been difficult to
carry out and the result of Luther and Towler was
possibly not final [5].
In the 90-s of the last century, a number of
laboratory experiments on measurement of the Newtonian gravitational constant were done with relative
uncertainties of about 100 ppm and less (Michaelis,
1995 [6]; Karagioz, 1996 [7]; Bagley, 1997 [8];
Fitzgerald, 1999 [9]; Jun Luo, 1999 [10]). These
results are partly summarized in Table. Nevertheless, the discrepancies between the values of the
gravitational constant obtained in these experiments
remained large enough. In particular, the value of
G = 6.7146 obtained in the Physikalish Technische
Bundesanstalt (Germany) [6], was more than by 40
standard deviations (i.e., more than 5000 ppm) above
the value of G recommended by CODATA in 1986.
As a result of such a scatter of G values, CODATA
had to increase the uncertainty significantly and in
1998 recommended the value G = (6.673 ± 0.010) ×
10−11 m3 kg−1 s−2 [11], with a relative error of
1500 ppm. i.e., the uncertainty of G knowledge
increased by almost a factor of ten!
128
In the following years (2000–2002), three new
results, with relative errors smaller than 50 ppm, were
published. These are the experiments of the University of Washington (USA), 2000, with a relative
error of 14 ppm [12], the University of Bermingham (Great Britain), 2001, with a relative error of
41 ppm [13], and the University of Zurich (Switzerland), 2002, with a relative error of 33 ppm [14].
Although the situation with G has been considerably
improved since the 1998 adjustment, these new results are not in complete agreement, as can be seen
from Table and Fig. 2. These new G values are not
crossed inside the confidence intervals. Based on
weighted means of results after 1998, all of which
round to G = 6.6742 × 10−11 m3 kg−1 s−2 , as well
as their uncertainties, the relatively poor agreement of
the data, and the historic and apparently continuing
difficulty of assigning an uncertainty to a measured
value of G that adequately reflects its true reliability,
CODATA has taken as the 2002 recommended value
G = (6.6742 ± 0.0010) × 10−11 m3 kg−1 s−2 with a
relative uncertainty of 150 ppm [15].
In 2003, a new result on G measurement was
reported by Armstrong and Fitzgerald from the Measurement Standards Laboratory (New Zealand),
GRAVITATION AND COSMOLOGY Vol. 14
No. 4
2008
STATUS OF THE EXPERIMENTS
371
G × 10–11, m3 kg–1 s–2
6.680
CODATA-98
6.678
[13]
6.676
[3] MSU
CODATA-02
[17]
[12]
[8]
6.674
[4]
6.672
[10] [14]
CODATA-86
[16]
CODATA-06
[7]
[2]
[9]
6.670
6.668
1970
1975
1980
1985
1990
Year
1995
2000
2005
2010
Fig. 2. The results of the world’s best experiments on G measurement and CODATA values. The dash line is the CODATA2006 value.
with an uncertainty of 40 ppm [16]. This uncertainty
is smaller by a factor of about 2.5 than the uncertainty of their previous result [9], but its standard
deviation interval does not intersect formally with the
CODATA-2002 one. Finally, the latest result of G
determination was published in 2006. It is the experiment of Schlamminger et al, from the University
of Zurich (Switzerland) [17]. This result, with an
uncertainty of 16 ppm, practically coincides with the
CODATA-2002 value.
The current value of the Newtonian gravitational
constant, recommended by CODATA in 2006, based
on all results available by the end of 2006, is G =
(6.67428 ± 0.00067) × 10−11 m3 kg−1 s−2 , with a
relative uncertainty of 100 ppm1 .
Looking at Fig. 2, we can conclude that some of
the values of the 30-year measurements of G do not
coincide with each other and with the latest CODATA
value. Until now there is no convincing explanation of such a large scatter in the G values obtained
in different experiments, although some hypotheses
have been discussed. One of them was expressed
by Kuroda [18]. The time-of-swing method of G
determination assumes that the elastic torsion constant Ke remains the same at each orientation of the
source masses, but this assumption has been put
to doubt. Kuroda has shown that damping of the
1
http://www.codata.org
GRAVITATION AND COSMOLOGY Vol. 14
No. 4
2008
torsion balance is caused by losses in the suspension
fiber (inelasticity of the fiber), and, for a Heyl-type
experiment, the measured value of G will be biased
upward by a factor of (1 + 1/πQ), where Q is the
quality factor of torsion oscillations.
Thus, the intricate situation which has developed
in the recent decades with G measurements makes
topical the performance of new experiments at a relative uncertainty of 10 to 30 ppm.
4. THE HISTORICAL EXPERIMENT
OF MOSCOW UNIVERSITY
The results of measurement of the Newtonian
gravitation constant at the Sternberg Astronomical
Institute of Moscow University has been published
in 1979 [3]. It used the time-of-swing method.
Changing of the gravitational configurations has been
achieved not by rotating the source masses but by
changing their position along the axis of the torsion
beam. The scheme of the experimental setup of
Moscow University is shown in Fig. 3.
Measurements of the gravitational constant have
been performed in four series of experiments. The first
series has been carried out from March to May 1975,
the second one from March to June 1976, the third
one from January to March 1977 and the fourth one
from October 1977 to January 1978. The first series
consisted of three measurements of G, the second and
372
MILYUKOV et al.
8
1
3
3
2
5
4
7
6
Fig. 3. The scheme of the experimental setup of SAI
MSU: (1) copper vacuum chamber; (2) torsion balance;
(3) source masses; (4, 5, 6) optical lever systems; (7) collimator for distance measurements; (8) vacuum pump.
cupied consecutively four fixed positions, therefore in
each such experiment it was possible to determine
three independent values of G.
The t-criterion has approved, that on a 95% significance level all four samples belong to one general
population with some average value. Therefore the
final result was found as an average in a sample containing 43 values of G and is equal to G = (6.6745 ±
0.0008) × 10−11 m3 kg−1 s−2 .
The t-criterion has also approved the assumption
that there is no time variation of G (the fourth series
of measurements was conducted 2.5 years after the
first series). On the basis of the measured value
of G, the masses and average density of the Earth,
the Moon and the Sun have been determined in the
metric system of units.
Finally, we would like to emphasize the following fact. The SAI MSU value, G = (6.6745 ±
0.0008) × 10−11 m3 kg−1 s−2 , was published in
1979. CODATA-1986 has recommended the value
of G = (6.67259 ± 0.00085) × 10−11 m3 kg−1 s−2 ,
which does not coincide with SAI MSU value. And
27 years after this experiment, in 2006, CODATA
recommended value of G which practically coincides
with this “old” value. From the position of modern experience, the SAI MSU value, due to the carefulness
of preparation and carrying out of the experiment,
despite the imperfection of measuring technology of
that time, has been apparently the least subject to
systematic errors. Thus, the 27-year-old value of G
is relevant and significant now.
5. TORSION BALANCE IS THE PRINCIPAL
TOOL OF THE EXPERIMENT
Fig. 4. Torsion balance No. 1, suspended in experimental
setup.
third ones of five measurements, and in the fourth
one ten measurements of G were made. To exclude
possible systematic errors in the experimental results,
between the series of experiments the torsion balance
was changed (two types of torsion balances were
used, Fig. 4) and repeated weighing of the source
masses and the torsion balance were conducted as
well as repeated measurements of the geometrical
dimensions of the torsion balance.
In each measurement of the first three series, one
value of the gravitational constant G was calculated.
In the last, fourth series, the attracting masses oc-
New technological approaches and optimization of
the configuration of experimental setups have shown
that the gravitational constant can be measured at an
accuracy level of 15 to 40 ppm (Table 1). Nevertheless, there are a number of problems which should
be solved in the experiments of this high level. One
of such problems is the precision of measuring the
period of eigen oscillations of the torsion balance. The
torsion balance is a complicated system with many
degrees of freedom, and due to nonlinear couplings
between them, new oscillations in so-called coupled
modes appear. This leads to perturbations of the basic
torsional mode and, as a consequence, to a great
uncertainty in determining its period.
Degrees of freedom of the torsion balance.
The detailed structure of the torsion balance which
is used in the Huazhong University of Science and
Technology (China) experiment on measurement of
G is shown in Fig. 5 [20]. The test mass, a rectangular
bar, is suspended from point O by a tungsten fiber of
GRAVITATION AND COSMOLOGY Vol. 14
No. 4
2008
STATUS OF THE EXPERIMENTS
373
Amplitude, × 10–8 rad
Z
Y
ζ(t)
Θ4
η(t)
ξ(t)
O
l
2
1
0
X
Θ5
–1
–2
Θ3
–3
Y1
Z1
0
2.5
5.0
Θ1
10.0
12.5 15.0
Time, min
Fig. 6. Swing oscillations on the ZY plane, excited by
seismic noise with an amplitude of about 1 mGal.
Θ2
X1
O1
7.5
ear equations, also containing fluctuating terms:
O2
l0
Θ̇1
(Jy + ml02 ) + mgl0 Θ1
τ1
= −mll0 Θ̈4 − ml0 ξ0 (t) + (Jy − Jx )Θ̈2 Θ3 ;
(Jy + ml02 )Θ̈1 +
Fig. 5. The torsion balance and the coordinate systems
are chosen to describe its motion.
(4)
Θ̇2
(Jx + ml02 ) + mgl0 Θ2
τ2
= −mll0 Θ̈5 − ml0 η̈0 (t) + (Jy − Jx )Θ̈1 Θ3 ;
(5)
(Jx + ml02 )Θ̈2 +
length l. To describe the dynamics of the torsion balance, we define two coordinate systems: a stationary
Cartesian coordinate system OXY Z, with the origin
at O located at the suspension point of the torsion
balance, and another coordinate system, O1 X1 Y1 Z1 ,
fixed rigidly with the test body—the torsion bar. Its
origin is at O1 the point of attachment of the fiber to
the torsion bar. The center of mass of the torsion balance is located at point O2 , at a distance l0 from point
O1 in the vertical direction. The suspension point
is driven by fluctuation forces (seismic noise), which
cause random displacements of the suspension point
ξ(t), η(t) and ζ(t), accordingly along X, Y and Z
directions. Other fluctuation forces, acting directly on
the torsion bar (e.g., random fluctuations of a residual
gas in the vacuum chamber, temperature variations
etc.), can provide the torsion balance rotation υ(t).
In such a setting of the problem, the system has
five degrees of freedom, to be designated as Θ1 , Θ2 ,
Θ3 , Θ4 , and Θ5 , respectively. Here the parameters
Θ1 and Θ4 represent rotation angles around the axes
Y1 and Y and describe swing oscillations in the plane
XZ. The parameters Θ2 and Θ5 represent rotation
angles around the axes X1 and X and describe swing
oscillations in the plane Y Z. The angle Θ3 represents
rotation about the Z-axis and describes the principal
torsion oscillations.
Hence, the set of equations (4)–(8), describing the
motion of the torsion balance, consists of five nonlinGRAVITATION AND COSMOLOGY Vol. 14
No. 4
2008
Jz Θ̇3
+ Ke Θ3 = (Jx − Jy )(Θ̇21 − Θ̇22 )Θ3
τ3
2
+ 2(Jx − Jy )Θ̇1 Θ̇2 Θ23 − (Jx − Jy )(Θ̇21 − Θ̇22 )Θ33
3
−2Jz Θ̈1 Θ2 + (Jx − Jy − 2Jz )Θ̇1 Θ̇2 + Ke ϑ(t); (6)
Jz Θ̈3 +
Θ̇4 2
ml + mglΘ4
τ4
= mll0 Θ̈1 − mlξ̈0 (t);
(7)
Θ̇5 2
ml + mglΘ5
τ5
= −mll0 Θ̈2 − mlη̈0 (t).
(8)
ml2 Θ̈4 +
ml2 Θ̈5 +
Here, m is the mass of the torsion balance; Jx , Jy ,
and Jz are the moments of inertia of the torsion bar
relative to the axes X1 , Y1 and Z1 , respectively; τi
are the time constants for each degree of freedom, Ke
is the torsion elastic constant, ϑ(t) is a fluctuation
function, describing random action on the torsion
degree of freedom.
Eqs. (4), (7) and (5), (8) describe swing oscillations in two perpendicular planes, ZX and ZY , while
Eq. (6) describes torsion oscillations. An analytic
solution of this set of equations was found in [19,
374
MILYUKOV et al.
Spectral amplitude
1
100
0.00174 Hz
10–2
10–4
2
34
5
6 78 9
10–6
10–8
10–10
10–4 10–3 10–2 10–1
100 101 102
Frequency, Hz
Fig. 7. Spectrum of torsion oscillations of an asymmetric
torsion balance. The peak at 1.74 × 10−3 Hz is an eigenmode of torsion oscillations, the peaks marked with the
indices 1 to 9 are coupled modes.
20]. Here, the set of equations has been solved by
a numerical method, and simulation of motion of the
torsion balance has been carried out.
Swing oscillations. We have studied the character of excitation of swing oscillations of the torsion
balance. A numerical experiment has shown that
oscillations in swing degrees of freedom are exited
by random noise of seismic origin and occur with a
time-variable magnitude (Fig. 6). A spectral analysis
of swing oscillations has shown that the oscillations
in each swing degree of freedom are beatings of all
quasi-harmonic swing modes with a random amplitude changing in time. The swing frequencies are
determined by the geometric parameters of the torsion
balance. The random nature of swing oscillations
is determined by the seismic noise that affects the
suspension point. It was also shown that even with
damping, due to the action of seismic noise, swing
oscillations are a steady process.
Torsion oscillations and mode couplings. Oscillations in the torsion degree of freedom are described by Eq. (6). The terms in the right-hand side of
this equation are nonlinear combinations of random
quasi-harmonic swing modes. So, the right-hand
side of this equation determines forced oscillations of
the torsion balance, which are torsional mode couplings, or coupled modes. The frequencies of the coupled modes are simply linear combinations of swingmode frequencies. The torsion balance represented
in Fig. 5 has an asymmetric configuration which is
characterized by inequality of all moments of inertia,
i.e., Jx = Jy = Jz . The swing-mode frequencies of
the asymmetrical torsion balance are also different.
Some of them can be close to each other. Their linear
combination will determine a low-frequency coupled
mode which can disturb the torsion mode.
A numerical simulation of the torsion balance motion with five degrees of freedom, described by the
set of equations (4)–(8), has been carried out. A
spectral analysis of the torsional degree of freedom
revealed a number of harmonics (Fig. 7). The most
intensive mode with the frequency 1.74 × 10−3 Hz is
the eigenmode of torsion oscillations. Other modes
are torsional mode couplings. Coupled modes with
the frequencies of 2 × 10−4 to 5 × 10−3 Hz are closest
to the eigenmode and can disturb the latter. Therefore the problem of high-precision measurements of
the torsional-mode frequency must be solved under
the condition of maximal suppression of the coupled
modes.
A traditional method of removing the effects of
low-frequency torsional mode couplings is to employ
a magnetic damper in the torsion system to overcome
the seismic noise and consequently to suppress the
intensity of swing modes. This method is called amplitude suppression and is used in the experimental
setup mentioned above.
The other way which we have proposed for suppression of combined modes is called frequency suppression. The main idea of this method is to choose
the geometry of the torsion balance in such a way as
to “move” the frequencies of the coupled modes to a
high-frequency range. In this case, the influence of
the coupled modes can be reduced to minimum. In
particular, a symmetric configuration of the torsion
balance (Jx = Jy = Jz ) leads to degeneracy of swing
modes, i.e., the swing frequencies in the ZX plane
coincide with those in the ZY plane.
CONCLUSIONS
There are a number of problems which must be
paid attention in order to achieve the desirable accuracy of G measurement. One of them is stability
of torsion oscillations. This stability also depends on
many factors, such as nonlinearity and thermoelasticity of the torsion wire, electrostatic effects, effects of
the geomagnetic field and so on. Among them there
is the effect of coupling of the swing modes to the
torsional mode. The torsion balance is a complicated
system with many degrees of freedom, and due to
nonlinear couplings between them, new oscillations
appear in the so-called coupled modes. This leads to
perturbations of the basic torsional mode, and, as a
consequence, to a big uncertainty in determining its
period. It was shown that this effect in the torsion
balance behavior arises directly from a driven combination of the swing modes of the balance, and the
swing modes are excited by the acting environmental
noise. To successfully suppress these types of mode
GRAVITATION AND COSMOLOGY Vol. 14
No. 4
2008
STATUS OF THE EXPERIMENTS
coupling in the torsion balance, amplitude or frequency suppression must be used in the experimental
setup [19, 20].
Measurement of fine effects in experimental
physics, especially in gravitational experiments,
needs high technology and nontrivial methods. Nevertheless, any experiment, even carefully prepared
and performed at the hottest features level, does
not give an absolute and final “knowledge”, but
only sets a new limit on a testable effect. From
this point of view, the technological development of
gravitational experiments, including measurement of
the Newtonian gravitational constant, is practically
an “infinite” process, at least in the nearest future.
We believe that new experiments, which are now in
progress in some world gravitational centers, give the
G value with a relative uncertainty within 10–20 ppm
and will be a successive stage in the knowledge of the
Newtonian gravitational constant.
ACKNOWLEDGMENTS
This work has been supported by the Russian
Foundation for Basic Research (No. 05-02-39014),
the National Basic Research Program of China
(No. 2003CB716300) and the National Science
Foundation of China (No. 10121503).
REFERENCES
1. P. R. Heyl, A redetermination of the constant of
gravitation, J. Res. Natl. Bur. Stand. 5 (256), 1243
(1930).
2. C. Pontikis, Détermination de la constante de
gravitation par la méthode de résonance, C. R.
Acad. Sci. Ser. B 274, 437 (1972).
3. M. U. Sagitov, V. K. Milyukov, et al., A new determination of the Cavendish gravitational constant,
Dokl. Akad. Nauk SSSR 245 (3), 567 (1979) [Sov.
Phys. Dokl. 245 (1–6), 20 (1981)].
4. G. G. Luther and W. R. Towler, Redetermination
of the Newtonian gravitational constant G, Phys.
Rev. Lett. 48, 121 (1982).
5. G. Gilles, The Newtonian constant, Metrologia 24
(Supplement), 56 (1987).
6. W. Michaelis et al., A new precise determination
of Newton’s gravitational constant, Metrologia 32,
267 (1996).
GRAVITATION AND COSMOLOGY Vol. 14
No. 4
2008
375
7. O. V. Karagioz and V. P. Izmailov, Measurement of
the gravitational constant with a torsion balance,
Izmer. Tekh. 39 (10), 3 (1996) [Meas. Tech. 39 (10),
979 (1996)].
8. C. H. Bagley and G. G. Luther, Preliminary results
of a determination of the Newtonian constant of
gravitation: A test of the Kuroda hypothesis, Phys.
Rev. Lett. 78 (16), 3047 (1997).
9. M. P. Fitzgerald and T. R. Armstrong, The measurement of G using the MSL torsion balance, Meas.
Sci. Technol. 10 (6), 439 (1999).
10. J. Luo et al., Determination of the Newtonian
gravitational constant G with a nonlinear fitting
method, Phys. Rev. D 59, 042001 (1999).
11. P. Mohr, Quantum electrodynamics and the fundamental constants, Advances in Quantum Chemistry
30, 77 (1998).
12. J. H. Gundlach and S. M. Merkowich, Measurement
of Newton’s constant using a torsion balance with
angular acceleration feedback, Phys. Rev. Lett. 85,
2869 (2000).
13. T. J. Quinn, C. C. Speake, S. J. Richmann, et al., A
new determination of G using two methods, Phys.
Rev. Lett. 87, 111101 (2001).
14. St. Schlamminger, E. Holzschuh, and W. Kundig,
Determination of the gravitational constant with
a beam balance, Phys. Rev. Lett. 89, 161102 (2002).
15. P. J. Mohr and B. N. Taylor, CODATA recommended
values of the fundamental physical constants:
2002, Rev. Mod. Phys. 77 (1), 1 (2005).
16. T. R. Armstrong and M. P. Fitzgerald, New measurement of G using the measurement laboratory
torsion balance, Phys. Rev. Lett. 91 (20), 201101
(2003).
17. St. Schlamminger et al., Measurement of Newton’s
gravitational constant, Phys. Rev. D 74, 082001
(2006).
18. K. Kuroda, Does the time-of-swing method give a
correct value of the Newtonian gravitational constant?, Phys. Rev. Lett. 75 (15), 2796 (1995).
19. V. K. Milyukov, The theory of motion of the torsion balance in inhomogeneous gravitational field
under the action of random noise. In: Problems
of Gravitation and Elementary Particle Theory,
(Energoizdat, Moscow, 1981), 12th issue, p. 128 (in
Russian).
20. X.-D. Fan et al., Coupled modes of the
torsion pendulum,
Phys. Lett. A, doi:
10.1016/j.physleta.2007.08.020 (2007).
Download